DISCLOSURE OF A DEPRESSIVE EPISODE DURING A JOB INTERVIEW

DISCLOSURE OF A DEPRESSIVE EPISODE DURING A JOB INTERVIEW The impact on perceivers cognitive, emotional and behavioural reactions 08-08-2006 Anouk Schippers, 273210 Erasmus University Rotterdam

DISCLOSURE OF A DEPRESSIVE EPISODE DURING A JOB INTERVIEW The impact on perceivers cognitive, emotional and behavioural reactions. Name: Anouk Schippers, 273210 Erasmus University Rotterdam Date: 08-08-2006 First supervisor: Dr. A.E.R. Bos Klinische psychologie (EUR) Second supervisor: Dr. E. Derous Arbeids- & Organisatiepsychologie (EUR)

CONTENTS Preface... 4 SUMMARY... 5 INTRODUCTION... 6 The social stigma of mental illness... 6 The process of stigmatization... 7 Functions of stigmatization... 9 The consequences of stigmatization... 10 Coping with social stigma... 13 Disclosure of mental illness on the workplace... 14 The present study... 15 METHOD... 18 Participants and design... 18 Procedure... 18 Measures... 19 RESULTS... 22 Manipulation checks... 22 Descriptive statistics... 22 Cognitive reactions... 24 Emotional reactions... 26 Behavioural reactions... 26 DISCUSSION... 28 Reason... 28 Seriousness... 30 Limitations of the present study... 31 Implications and future research... 32 LITERATURE... 34 APPENDIX A. Scenario (depression/high seriousness)... 38 APPENDIX B. Questionnaire 1... 46 APPENDIX C. Questionnaire 2... 52 3

Preface Over the past nine months I have been working on my master thesis to obtain the degree Master of Science in Clinical and Health Psychology at the Erasmus University Rotterdam. The subject of this thesis is stigmatization of psychiatric patients. I have been investigating the role of earlier depressive complaints during job applications. This document presents a detailed description of the process and the results of this study. First and foremost I would like to thank my supervisor Arjan Bos. He constantly supported me during the past nine months and was always willing to help me whenever I needed help. Thanks to his advice and critical comments I can finish my master Clinical and Health Psychology successfully. Furthermore, I would like to thank my second supervisor Eva Derous for her valuable suggestions and feedback during the process. Last but not least I want to give special thanks to my boyfriend, brother and parents for their love and support during the past four years I was a student at the Erasmus University Rotterdam. Rotterdam, The Netherlands 7 August 2006 Anouk Schippers 4

SUMMARY Earlier research provides evidence for the existence of stigmatization in important life domains of psychiatric patients. The present study is a scenario study which investigates the role of earlier depressive complaints in job applications. The impact of seriousness of complaints is also investigated. Economy and business administration students (N = 180) were given a scenario. The scenario provided a detailed description of an application procedure. A 33-year old woman applied for the job. She perfectly matches the requirements of the job but she has a career gap in her resume. The reason for this career gap (depression, car-accident or no information) and the seriousness of the complaints (low seriousness vs. high seriousness) were manipulated. Differences in perceivers cognitive, emotional and behavioural reactions towards the applicant were measured. Results showed that disclosure of earlier depressive complaints during a job interview has a negative influence on the chance that the perceiver will hire the applicant and on the personal assessment of the applicant. More negative personal characteristics were attributed towards the depressive applicant compared to the applicants in the other two conditions. Implications of the results and recommendations for future research are discussed. 5

INTRODUCTION The social stigma of mental illness Originally the term stigma refers to bodily signs made by the ancient Greeks (Goffman, 1963). These signs were burned or cut in the body of a person and showed that the person had a bad or lower status, like slaves or criminals. Nowadays there is no widely accepted definition of stigma. Goffman (1963) defined stigma as an attribute that is deeply discrediting (p. 3). Jones and colleagues (1984) agreed with this definition, but place stigma in a relational perspective. A condition labeled as deviant for one person, may be attractive according to another individual. They stated that a specific attribute (mark) of a person links a person to an undesirable characteristic (stereotype). According to the relational perspective, this link can be different across cultures. For example, labeling a person as mentally ill will have different effects in different cultures (Angermeyer, Buyantugs, Kenzine, & Matschinger, 2004). Angermeyer and colleagues (2004) showed that labeling a person as schizophrenic will lead to perceptions of dangerousness in Germany. In contrast, no perceptions of dangerousness were found in Russia and Mongolia. What seems discrediting in one culture can be the norm in another. According to the definition of Link and Phelan (2001) stigma only exists when elements of labeling, stereotyping, separating, status loss and discrimination co-occur in a power situation that allows these processes to unfold. People with a mental illness belong to one of the most stigmatized groups in our society. Research on public opinions in both The Netherlands and Germany shows that people with schizophrenia are seen as dangerous and unpredictable (Angermeyer & Matschinger, 2005; Boon, Nugter & Dijker, 2004). Although people with alcohol dependence and schizophrenia were mostly rejected, people with depression experience a substantial amount of stigma and discrimination too (Angermeyer & Matschinger, 1997). The latter study showed that 45% of the general population would not recommend someone with a major depression for a job and 26% would not like to have a depressed person as a neighbour. These results contradict Western legislation 6

which is prescribing equal opportunities in work and education, regardless of ethnic background, social class, age, political preference, religious conviction or sexual orientation. Despite this legislation many minority groups, such as mentally ill patients, do not have equal opportunities compared to the general population (Angermeyer & Matschinger, 1997; Boon et al., 2004; Corrigan et al., 2000; Ellemers & Barreto, 2006; Markowitz, 1998). According to the statistics of the UWV (a Dutch organization for employee-insurances) 17.064 employees have applied for invalidity insurance in the period of January till September 2005 (UWV, 2006). Almost 30% (4.916 employees) did this because of mental complaints. The purpose of the present study is to explore the differences in perceivers cognitive, emotional and behavioural reactions towards an applicant with an earlier depression compared to a car-accident and a control condition. Moreover, the study investigates the impact of seriousness of depressive complaints on these same emotional and behavioural reactions. The process of stigmatization In 1988 Weiner and colleagues proposed an attribution-emotion model of social stigma (Weiner, Perry & Magnusson, 1988). According to this theory there is a causal relationship between cognitions and specific affective reactions towards a stigmatized person. Their attribution-emotion model states that stigmatization is caused by two affective reactions: anger and pity. These emotions are caused by the perceived controllability of the stigmatized condition. When personal controllability is high, people will respond with anger and stigmatization towards a (mentally) ill person. Additionally when personal controllability is low, people will respond with pity instead of anger. For example, according to this theory a perceiver will respond with more anger towards a psychotic patient when psychotic complaints are caused by the use of cannabis. This same perceiver will respond with more pity when psychotic complaints are caused by hereditary factors. Dijker and Koomen (1996, 2003) extended this attribution theory with other cognitions and emotions. They claimed that, besides pity and anger, the degree in 7

which a person breaks social norms, the degree in which a person suffers, and contagiousness of the deviant condition play a significant role in the stigmatization process too. In addition, these cognitive dimensions underlie more then the two affective states proposed by Weiner (1988). Dijker and Koomen (1996, 2003) extended these cognitive states with gloating and fear which, according to their theory both lead to social rejection (see Figure 1). Contagiousness Seriousness Personal responsibility Norm violation + + Fear + + - Pity - + + + - Anger Social rejection of psychiatric patients Figure 1. A cognitive-emotional model of stigmatization of psychiatric clients, based on Weiner et al. (1988); Dijker and Koomen (1996) and Dijker and Koomen (2003). An important question is to what extent these models can be applied to stigmatizing responses towards people with a mental illness. Results from research on this topic supported the attribution theory proposed by Weiner (1988) and suggested that mental health disabilities were rated more negatively on the factors controllability and stability compared to physical disabilities (Corrigan, River, Lundin, Uphoff Wasowski, Campion, Mathisen, Goldstein, Bergman & Gagnon, 2000). Another study of Corrigan and colleagues (2003) showed that, consistent with the theories of Weiner (1988) and Dijker and Koomen (1996, 2003), the emotions pity, anger and fear were affected by beliefs about the controllability of the mental illness and lead, in the high 8

controllability cases, to rejecting responses (Corrigan, Markowitz, Watson, Rowan & Kubiak, 2003). This study also found that perceived dangerousness of a psychiatric patient increased fear which in turn leads to rejection and avoiding behaviour (Corrigan et al., 2003). Finally, this study also revealed that familiarity with mental illness had an influence on discriminatory and emotional reactions. Participants who were more familiar with mental illness were more likely to offer interpersonal help and less likely to avoid mentally ill patients (Corrigan et al., 2003; Angermeyer & Matschinger, 1997; Boon et al., 2004). Angermeyer and Matschinger (1997) attributed this result to an increase in positive emotional reactions and reduced feelings of anxiety towards mentally ill people. An experimental study of De Graaff-Wijnberg and colleagues (2006) found some support for the theory of Dijker and Koomen (1996, 2003) as well. In their study a scenario was given to 320 health service workers. This scenario described a colleague who returned to work after a period of absence due to depressive complaints. Personal responsibility, seriousness of complaints, and openness about the complaints were manipulated. Evaluations of personal qualities and emotional reactions toward the person were measured. Results showed that participants reported more fear when the colleague had serious versus less serious complaints. However, no support was found for the impact of personal responsibility on stigmatization (De Graaff-Wijnberg et al., 2006). Functions of stigmatization There are some theories about the functions of stigmatization as well (Bos, 2001; Crocker, Major, & Steele, 1998; Dovidio, Major & Crocker, 2000). Stigmatization can enhance control over threatening situations, enhance self-esteem, cause in-group enhancement, or buffer against anxiety. A stigma provides information about possible danger for a perceiver. Knowledge about this possible danger enhances control over threatening situations. For example, HIV related stigma provides information about the contagiousness of the disease (Bos, 2001). Knowledge about this contagious aspect of the disease makes that 9

a perceiver can protect itself by avoiding HIV patients. The same applies to the stigmatization of psychiatric patients. Several studies showed that 50% of the respondents thought that psychiatric patients can be aggressive (Boon et al., 2004; Angermeyer, Beck & Matschinger, 2003, Angermeyer et al., 2004). By stigmatizing psychiatric patients through, for example, avoiding them, perceivers can protect themselves against the aggressiveness of these patients. Secondly, stigma can enhance self-esteem through downward comparison (Crocker et al., 1998; Dovidio et al., 2000, Bos, 2001). Downward comparison means comparing oneself with a less fortunate person. This comparison can enhance one s self-esteem, because of the positive influence on the person s sense of well-being. Through active stigmatization of a psychiatric patient one can enhance their own selfesteem by making the other person inferior. Furthermore, stigmatization can cause in-group enhancement as well (Crocker et al., 1998; Dovidio et al., 2000, Bos, 2001). Stigmatization of a specific group increases the superiority of the stigmatizing group and the differences between the groups. In this manner, the stigmatizing group maintains a positive social identity. Finally, stigmatization can buffer against existential anxiety (Crocker et al., 1998; Dovidio et al., 2000, Bos, 2001). Awareness of your own mortality and vulnerability can generate feelings of anxiety in a person. For example, contact with a HIV patient reminds someone about their own mortality which in turn increases their anxiety (Bos, 2001; Dechesne, Janssen, Van Knippenberg, 2000). The consequences of stigmatization Research among people with mental illnesses has demonstrated that stigmatization has a negative impact on stigmatized people (Link, Struening, Neese-Todd, Asmussen & Phelan, 2001; Hayward, Wong, Bright & Lam, 2002; Rosenfield, 1997; Markowitz, 1998; Major & O Brien, 2005). The initial labeling theory states that when a person is labeled as mentally ill by a professional, this label will have negative consequences for the mentally ill person. The label will lead to rejection and discrimination in society, which in turn will have a negative influence on outcomes for the mentally ill. Because 10

of the enormous and sustained criticism on the initial labeling theory, a modified version of this theory is proposed (Link, Cullen, Struening, Shrout & Dohrenwend, 1989). This modified version replaced the strong claim of the initial labeling theory, that a person with a mental illness label is doomed, with a more adapted point of view. Link and colleagues (1989) proposed that, through socialization, every person develops a set of stereotypical attitudes about mentally ill patients and about how most people treat these patients. When psychiatric patients are labeled as mentally ill by professionals, these attitudes become personally relevant. These patients expect to be treated according to their own stereotypical attitudes and these attitudes become self fulfilling prophecies. When psychiatric patients expect to be devalued and discriminated against, they will use certain coping strategies, such as withdrawal. These ineffective coping strategies will lead to a decrease in social networks, lower self-esteem and less job opportunities. To summarize, the modified labeling approach claims that, although stigmatization can t directly cause a mental disorder, it can have a negative influence on the course of the illness. It causes a vulnerability to new disorders or a repetition of the present episode (Link et al., 1989). Critics of the labeling theory argue that labeling has positive consequences for the mentally ill person as well. Rosenfield (1997) for example, tested a theoretical framework that stated that labeling would lead to negative ánd positive consequences. She found that positive consequences of labeling, such as receiving treatment and high quality services, have a positive influence on life satisfaction. Negative consequences on the other hand, such as stigmatization and discrimination, have a negative influence on life satisfaction. Furthermore, these two influences are related to subjective quality of life through their associations with the person s self-concept. The feeling of being stigmatized affects the person s self-concept and self-efficacy, which in turn has a negative influence on the quality of life. The received treatment and high quality services, on the other hand, have a positive influence on the quality of life, because it increases people s self-concept en self-efficacy. Additionally, much research has been carried out regarding the adverse effects of stigma. Most of these research states that stigmatization has a negative influence on a persons self-esteem (Link et al., 2001; Vinken, Bos, Bolman, Van der Plas, 2005; 11

Hayward et al., 2002; Markowitz, 1998). Link and colleagues (2001) measured perceived stigmatization and self-esteem among 70 members of a clubhouse program for people with mental illness on different time points. Even after adjusting for baseline self-esteem and depressive symptoms, perceived stigmatization strongly predicted self-esteem. Hayward and colleagues (2002) demonstrated this negative effect of stigma on the self-esteem of people with manic depression. Unfortunately, this study of Hayward and colleagues (2002) concerned a cross-sectional design. For this reason, an important limitation of this study is the possible bi-directional relationship between stigma and self-esteem. Stigmatization may lead to lower selfesteem, but those with lower self-esteem, may be more vulnerable for stigmatization as well (Hayward et al., 2002; Markowitz, 1998). Furthermore, research showed that the relationship between stigmatization and self-esteem is dependent on the diagnosed disorder. This relationship is especially apparent in depressive-anxiety types of disorders and not in psychotic disorders like schizophrenia (Markowitz, 1998; Vinken et al., 2005). Finally, Major and O Brien (2005) stated that stigma has a direct and an indirect effect on the stigmatized person. It directly affects a stigmatized person through mechanisms as negative treatment and discrimination, expectancy confirmation and automatic stereotyping. Negative treatment and discrimination has a direct influence on the access to important life domains for the stigmatized person, such as opportunities for finding a job and living space (Angermeyer & Matschinger, 1997; Boon et al., 2004; Markowitz, 1998). For example, 38.4% of the German public would rather not rent a room to someone with a major depression (Angermeyer & Matschinger, 1997). A Dutch questionnaire among neighbours of a psychiatric clinic showed that only 52% of the respondents would accept a psychiatric patient as a neighbour. Only half of the respondents had a positive attitude towards psychiatric patients (Boon et al., 2004). According to expectancy confirmation process, also referred to as self-fulfilling prophecies, a stigmatized person will behave according to the negative stereotype. This behaviour confirms the expectations and even leads to changes in self-perception. In addition, stigma also affects the stigmatized person via indirect processes as threats to personal and social identity. According to this theory 12

stigmatization of a psychiatric patient puts this patient at risk of experiencing threats to his or her social identity (Major & O Brien, 2005; Crocker & Major, 1989). Coping with social stigma Being stigmatized can be a very stressful experience and can have a lot of negative consequences for the stigmatized individual. Important to know is if stigmatized individuals can cope with these stressors and which strategies work best. Different kinds of coping strategies can be found in the empirical literature, such as avoidance, denial and acceptance (Crocker et al., 1998; Link, Mirotznik and Cullen, 1991; Miller & Major, 2000; Vinken et al., 2005). According to Miller and Major (2000) the effectiveness of a coping strategy depends on the kind of person, the situation and on the person s goal. In other words, different strategies are effective for different persons in different situations. Miller and Major (2000) differentiate between problem-focused coping and emotion-focused coping. Problem-focused coping is effective when a person s goal is reducing the stressor, such as trying to conceal or reduce the stigmatizing attribute. When a person s goal is controlling the stressful emotions caused by stigmatization through minimizing negative emotions, emotion-focused coping seems to be more effective. An example of emotion-focused coping is downward social comparison (Bos, 2001; Crocker et al., 1998; Miller & Major, 2000). A stigmatized mental patient can use this social comparison strategy to protect one s self-esteem against the negative consequences of stigmatization. Earlier research shows that downward social comparison plays a moderating role in the relationship between stigmatization and self-esteem (Crocker et al., 1998; Vinken et al., 2005). Crocker, Major and Steele (1998) differentiate between three different coping strategies which stigmatized people use to avoid negative outcomes. The first strategy is attributing negative outcomes to external factors, such as prejudice and discrimination (Crock et al., 1998). A second strategy which stigmatized persons can use is the downward social comparison technique (Bos, 2001; Crocker et al., 1998; Miller & Major, 2000). The last strategy is making the stigmatized aspect less 13

important for the self-esteem by psychological disengaging the self-esteem from possible negative outcomes in a specific domain (Crocker et al., 1998). Failure in a specific domain caused by stigmatization, for example social interaction, can have negative consequences for the self-esteem. By psychological disconnecting the selfesteem from this domain, people can avoid this negative impact on the self-esteem. Disclosure of mental illness on the workplace Employees with a concealable stigma such as mental illness can choose between concealment or disclosure of their condition. Concealment will probably lead to less stigmatizing reactions, because co-workers are unaware of the stigmatizing attribute of their colleague. On the other hand, disclosure can lead to social support of colleagues. According to Goldberg and colleagues (2005), many professionals in the American public mental health field recommend their mentally ill clients to disclose their illness when applying for a job. Reasons for this preference for disclosure include the opportunity to claim rights for people with a (psychiatric) disability, such as adapted working hours or a low-stress job, and the belief held by many mental health professionals that employees with psychiatric complaints will chronically experience negative symptoms (Goldberg, 2005). Moreover, sometimes it isn t even possible to conceal a mental illness during a job interview, because many psychiatric patients are accompanied by a rehabilitation worker when applying for a job. Only a few studies tried to explore the consequences of disclosing a mental illness (Corrigan, 2005; De Graaff-Wijnberg et al., 2006; Goldberg et al., 2005; Quinn, Kyoung Kahng & Crocker, 2004). These studies not always support this recommendation of American professionals. Quin and colleagues (2004) studied the effects of disclosing mental illness on test performance. Results indicated that psychiatric patients who directly disclosed their mental illness before taking a test, performed inferior compared to patients who did not disclosed their mental illness (Quin et al., 2004). In a study of De Graaff-Wijnberg and colleagues (2006), openness about depressive symptoms on the workplace was associated with more feelings of fear, pity and anger in coworkers (De Graaff-Wijnberg et al., 2006). This result shows 14

that disclosure of a mental illness in the workplace will lead to positive reactions, like pity, but to negative emotions, like fear and anger, as well. Ellemers and Barreto (2006) reviewed empirical research concerning the effects of hiding a devalued identity in the workplace. Their review concludes that, although concealment will protect an employee against stigmatizing reactions, it has negative effects as well. For example, hiding a stigmatizing attribute brings substantial emotional and cognitive costs along. These costs can be so severe that they will increase instead of decrease the stress on the workplace. Consequently, people who are concealing their mental illness on their work can experience such an amount of stress that they will perform suboptimal (Ellemers & Barreto, 2006). Corrigan (2005) discriminates between different levels of disclosing, such as secrecy, selective disclosure, indiscriminant disclosure and broadcasting. In case of the first two types of disclosure, still some secrecy exists. Secrecy means total concealment of the mental illness and selective disclosure means disclosing a mental illness to specific persons, such as the supervisor on the workplace. Jones (1984) states that concealment is mainly an effective approach in short-term contact and casual interactions. When a person chooses for indiscriminant disclosure, everyone could find out about their mental illness, because no efforts are made to conceal their mental illness. This does not mean someone is continuously telling everyone about their illness. This is the case when a person choose to broadcast their mental illness. Broadcasting means educating people about mental illness by telling one s story. According to Corrigan (2005) broadcasting is not only beneficial because of a lack of secrecy. Broadcasting also fosters the power over the experience of mental illness and stigma (Corrigan, 2005). Until now only little research has been done concerning the pros and cons of disclosing a mental illness during a job interview. The present study Only little research has been done regarding the influence of disclosure of depressive symptoms on cognitive, emotional en behavioural responses of perceivers and the 15

influence of seriousness of depressive complaints on these reactions. The study of De Graaff-Wijnberg and colleagues (2006) showed that serious complaints increase fear in perceivers and made them attribute more negative personal traits towards a depressed person, compared to less serious complaints. Moreover, serious complaints had no effect on the amount of compassion. The present study focuses on disclosure of depressive complaints in a job interview and examines the influence of seriousness of complaints on cognitive, emotional and behavioural reactions in perceivers. Most of the research on stigmatization has been aimed at the stigmatized person. The present study also takes some moderating factors of the stigmatizing person into account. Reason We expect that the reason for the career gap will have a significant effect on the cognitive reactions in perceivers. Perceivers will attribute more negative personal skills and traits towards the applicant in the depressive condition compared to the applicant in the car-accident and control condition (hypothesis 1). Furthermore, we expect that the applicant in the depressive condition will evoke more negative emotional reactions in perceivers compared to the applicant in the car-accident and control condition (hypothesis 2). Finally, we expect that the depressive applicant will evoke more avoiding behaviour and less approaching behaviour compared to the applicant in the car-accident and the control condition (hypothesis 3). Avoiding behaviour is behaviour that creates social distance between the perceiver en the applicant. Approaching behaviour is the opposite, such as helping someone. Seriousness We expect that seriousness of complaints will have a significant negative effect on the cognitive reactions in perceivers. Perceivers will attribute more negative and less positive personal skills and traits towards the applicant in the high seriousness condition compared to the applicant in the low seriousness condition (hypothesis 4). Based on the model of Dijker and Koomen (1996, 2003) we expect that the applicant in the high seriousness condition will evoke more negative emotional reactions in perceivers compared to the applicant in the low seriousness condition (hypothesis 5). 16

Finally, we expect that the applicant in the high seriousness condition will evoke more avoiding behaviour and less approaching behaviour compared to the applicant in the low seriousness condition (hypothesis 6). 17

METHOD Participants and design Participants (N=180, 90 males and 90 females) were economy (N=23) or (International) Business Administration (N=157) students. Manipulations failed for 11 participants. These participants were excluded for further analyses (N = 169, 81 males and 88 females). The mean age of the sample was 20.62 years (SD = 1.95). Participants were randomly assigned (in order of entry) to a 3 (reason: depression, car-accident, no information) x 2 (seriousness: high vs. low) between-subjects design. Procedure Participants were contacted by telephone and were asked to join the experiment. An appointment was made in the Erasmus Behavioural Lab. The experimenter welcomed the participants and explained the procedure of the study. The experimenter explained that every participant would receive a scenario about an application procedure and several questions about this procedure. It was pointed out that participants had to read the scenario carefully and had to imagine being a recruiter. Subsequently, one of the six scenarios was presented. Participants read the scenario and filled out the questionnaire. At the end of the experiment participants were debriefed and paid for their participation (10 euro). Scenario Scenarios were provided in booklets to participants. Each booklet provided a detailed description of one scenario (i.e., application procedure). Participants were asked to imagine being a member of an application committee. The company of this committee was looking for an administrative employee and the job advertisement was presented. This advertisement gave a description of the job, the required skills and the necessary personal skills. These personal skills were flexibility, creativity, independence, exactitude and cooperativeness. Next, participants had to read an application letter of 18

a 33-year old woman who was interested in the job. The curriculum vitae of this woman perfectly matched the description in the job opening, but had one particular detail. Five years ago, the applicant had a career gap in her resume. Since the gap she had well functioned in a similar job. The applicant was invited for a meeting with the application commission and a description of this meeting was presented. During the meeting the applicant was asked why she had this career gap in her curriculum vitae. The reason for the career gap in the curriculum vitae was manipulated. The participant had either suffered from a depression, or had been involved in a car accident and needed a period of time to recover from her injuries, or the career gap was noticed but the committee did not ask about the reason. Seriousness was manipulated through varying the length of the career gap. In the high seriousness condition the career gap was one year, whereas this was two months in the low seriousness condition. Measures After reading the vignette, participants answered several questions concerning their cognitive, emotional and behavioural reactions towards the applicant. Answers were measured on a 7-point Likert scale (1 = not at all, 7 = very much). In the end, background questions were asked concerning gender, age, education and familiarity with depressive symptoms. Cognitive reactions Subsequently, questions were asked about how well she matched with the required personal skills described in the job opening (flexibility, creativity, independence, cooperative and exactitude) and about different traits (e.g. friendly, sensitive, shy, insecure, introvert). These different traits were subdivided into categories; positive and negative traits. Positive traits were friendly, nice and talkative (Cronbach s alpha =.75) and negative traits were introvert, shy, stretched, insecure, hesitative and sensitive (Cronbach s alpha =.81) 19

Emotional reactions Emotional reactions were measured with the question: When I think about Linda (the applicant), I feel The items were based on previous research of Weiner, Perry and Magnusson (1988) and Dijker and Koomen (2003) and concerned pity, anger and fear. Pity was measured with the items compassion, involvement, pity and sympathy (Cronbach s alpha =.64). Anger was measured with the items annoyance, fury, irritation and anger (Cronbach s alpha =.87). Fear was measured with the items nervousness, tenseness, uncertainty and fear (Cronbach s Alpha =.84). Behavioural reactions Additionally, intentional behaviour was measured with the questions: When I think about Linda, then and When I think about Linda, I would These reactions were measured with ten items about social interaction. Five items described approaching behaviour, such as sitting next to Linda during lunch and asking Linda for advice when you have a problem (Cronbach s alpha =.83). The other five items described avoiding behaviour such as keeping some distance between Linda and yourself and avoiding Linda (Cronbach s alpha =.91). Further, some questions concerning the intentions to select the applicant and her aptitude for the job were asked (e.g., How big is the chance that you will select Linda? and How suitable is Linda for the job?). Control variable During all analyses we controlled for participants motivation to control prejudice. A Dutch translation of the 12-item Motivation to Control Prejudice Against Persons With Aids (PWAs) Questionnaire was used (Dunton & Fazio, 1997; Pryor, Reeder, & Landau, 1999). Because the present study is about psychiatric patients we changed persons with aids to persons with a mental illness in the questionnaire. The questionnaire contains 12 items. In the present study both alphas were.72 and.76. 20